In a 5-year longitudinal study, social support was investigated as a predictor of low back pain (LBP) in 46 initially asymptomatic individuals. Distinguishing between colleagues in general and the colleague one feels closest to, the authors analyzed constellations of support from supervisor (which had positive effects) and closest colleague (which had detrimental effects). Configural frequency analysis yielded a type characterized by high support from one's closest colleague and low support from one's supervisor at Time 1 and more LBP and disability at Time 2. Controlling for negative affectivity did not change findings. Results are interpreted in terms of being dependent on that one source of support in an unsupportive environment, which creates feelings of dependence, incompetence, and reciprocity obligations.